February 16, 2026

Deploying Agentic AI at Scale: Infrastructure, Reliability, and Risk with Ran Aroussi

Deploying Agentic AI at Scale: Infrastructure, Reliability, and Risk with Ran Aroussi

Listen

Deploying Agentic AI at Scale: Infrastructure, Reliability, and Risk with Ran Aroussi
0:00 / --:--
Business of Tech: Daily 10-Minute IT Services Insights

Show notes

Agentic AI is being deployed as production infrastructure in enterprise settings, but prevailing frameworks remain unreliable for mission-critical operations. Dave Sobel and Ron Aroussi from Muxie underscored that while AI agents are functional—especially in non-deterministic contexts like customer support—expectations of deterministic, workflow-based reliability are not met. The move from demonstration agents to production-scale tools brings heightened attention to issues of reliability, observability, and especially risk of vendor lock-in for Managed Service Providers (MSPs) and their clients.

Operational deployment of AI agents currently gravitates toward roles with minimal operational risk, such as customer-facing chatbots or internal chief-of-staff assistants. Aroussi explained that while such agents can automate initial support tiers and internal daily briefings, their unpredictability and potential for error limit their use in processes demanding strict oversight and accountability. He identified two core use cases—external (customer support) and internal (personalized information management)—explicitly noting that agents are best positioned to augment rather than fully automate complex workflows at this stage.

A critical risk for MSPs lies in attempting to retrofit existing software frameworks to support agents, which introduces integration complexity and increases the likelihood of operational failures. Purpose-built infrastructure for agentic AI offers better alignment between AI capabilities and production requirements, with Aroussi citing drastically reduced hallucination rates and improved oversight when using native tools. Open source is identified as a foundational element for AI development, but it incurs its own risks, particularly around third-party code quality and the long-term sustainability of community-driven projects.

The practical implication for MSPs and IT service providers is clear: a cautious, incremental adoption approach focused on low-risk use cases, coupled with rigorous controls on agent permissions and robust audit trails, is essential. Decision-makers should avoid assuming agents operate with the reliability or accountability of traditional software, prioritize operational transparency, and ensure that responsibilities for agent actions are clearly defined and enforced at the implementation level. Vendor lock-in and software provenance remain significant governance concerns as agentic AI moves from experiment to infrastructure.

About Business of Tech: The daily MSP news and analysis podcast for managed service providers and IT services leaders. Host Dave Sobel covers AI for MSPs, MSP pricing and profitability, MSP M&A and private equity, vendor partner programs, MSP cybersecurity and liability, and MSP operations — and answers the question every story raises: Why do we care? Produced by MSP Radio.

💼 All Our SponsorsMSP Radio is supported by our partners:

ABC Solutions · CometBackup · Firetail · Guardz · HaloPSA · LogMeIn · Mailprotector · Oveka · OpenText · Pax8 · Rythmz · ScalePad · TimeZest

Supporting the IT services community through insights, analysis, and transparency.

🚀 Join Business of Tech PlusGet exclusive access to investigative reports, vendor analysis, leadership briefings, and more.

🎧 Subscribe to the Business of TechWant the show on your favorite podcast app or prefer the written versions of each story?

📰 Story Links & SourcesLooking for the links from today’s stories?

Every episode script — with full source links — is posted at:

🎙 Want to Be a Guest?Pitch your story or appear on Business of Tech: Daily MSP News & IT Services Analysis

🔗 Follow Business of Tech

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

About Ran Aroussi

Ran Aroussi, podcast guest

Founder of VarOps • Building software and Resident AI into companies that can't afford to guess.

35+ years Production coding experience30M+ Open-source downloads per month50K+ GitHub stars3B+ Ads delivered daily by systems he built

Ran Aroussi has spent 35 years building software infrastructure. Now he thinks we’re building AI for businesses backwards.

Most companies are adding copilots, chatbots, and agents, then asking employees to learn how to use them. Ran’s argument is almost the opposite: people shouldn’t have to adapt to AI. AI should adapt to the company.

He’s the founder of VarOps, where he’s building what he calls "Resident AI": AI that lives inside an organization, learns how it actually operates, and works through the tools and workflows people already use. No new destination. No constant prompting. Ideally, employees barely notice it’s there.

The idea grew out of a problem Ran believes the AI industry has underestimated: AI doesn’t understand organizations.

A chatbot might have access to every document, meeting, and message in a company and still not understand why a decision was made, which unwritten rule matters, who actually knows how something works, or that the official process hasn’t been followed in three years.

Retrieval gives AI information. It doesn’t necessarily give it understanding.

Ran’s work focuses on building that missing layer: a living model of the organization itself – its knowledge, decisions, relationships, processes, and unwritten operating context. He describes it as an "Organizational Language Model" (OLM) rather than another LLM with access to company data.

View profile →

Podcast Hosts: Get Ran’s Media Profile

Click to Download